Today’s security professionals should consider using artificial intelligence (AI) when it comes to security issues such as responding to alarms and identifying people on video camera footage.
The Security Industry Association’s (SIA) AI, Drones, and Robotics Interest Group recently hosted the first part of an AI virtual town hall event featuring four professionals in the security industry. Additional virtual events are anticipated in the future.
Steve Reinharz, chairman of the AI, Drones, and Robotics Interest Group; founder and CEO of Artificial Intelligence Technology Solutions; and President of Robotics Assistance Devices, said the group “helps create standards, helps advocate for proper use of AI, and helps share information in order to strengthen SIA’s membership for the use of AI applications because this group really feels AI-infused everything is inevitable.”
What Is AI?
AI includes symbolic learning, which trains computers to learn the same way as a human brain and can include robotics and computer vision such as industrial automation, and machine learning, which allows AI to imitate intelligent human behavior. Machine learning can include deep learning, which processes sensory data, including object recognition, and is useful in the security industry, as well as statistical learning, including speech recognition and language processing.
Robotics is the driving application for AI. Robots navigate their space, look for abnormalities, and act. As Travis Deyle, Cofounder and CEO of Cobalt Robotics, explained, “Increasingly as these edge devices become more and more capable, as they have mobility and manipulation, you will see robots going out and doing these basic tasks that today we relied exclusively on people because it was our only option.”
Basic tasks security managers could use robots for include:
- Intruder alerts
- Alarm response
- Leaks and spills
- Exterior open doors
- Overflowing trash cans
- Virtual reception
AI Video Security Platform
Security professionals often have a hard time finding certain events when combing through their security footage after an incident, Ron Rothman, President of Turing AI, noted. To make this easier, AI can be used in video surveillance platforms to search the footage.
AI can comb through footage specifically to:
- Determine if and how emergency exits are blocked and send an alert to prevent businesses from getting fined.
- Identity specific people based on attributes like clothing.
- Locate a vehicle of interest based on color and model.
- Get an immediate alert if a person or vehicle comes on-site.
- Easily share these video clips with investigative personnel.
AI and Guarding Companies
Security guarding companies need help meeting their clients’ needs, but regular security cameras often don’t provide enough protection.
According to Reinharz, AI can help security guarding companies in the following ways:
- Trespassing detection—Intelligently determine when people are trespassing to combat more serious crimes such as rioting, vandalism, and theft.
- Loitering detection—Track individuals to determine how long they remain in a location to prevent people from sleeping overnight in unauthorized places.
- Firearm detection—Help detect individuals who have weapons, and provide additional safety to those on the property, such as students at a school.
- Periphery patrols—Allow a mobile device to help secure the perimeter of a property.
Security managers, often called “customers” by security companies, must decide how best to use their resources and allocate budgets, and they often have difficult decisions to make when determining how to provide the best security while also watching the bottom line.
According to Reinharz, AI devices can “help save customers money, help guarding companies reduce their reliance on constantly hiring additional humans and the end result is hopefully to be able to rebalance end users spending so that the pay rate for guards can be higher, it can be a real career. If we don’t have the turnover, we can provide better security with better professional staff that are going to get paid more when they are teamed with autonomous remote services.”
While the rules of facial recognition can vary by state and by city, speakers at the town hall event said their products require security managers to “opt in”; however, most choose not to because of ethical and racial concerns but admit facial recognition is sometimes used for access control, comparing it more to the face ID on smartphones. But, there are other ways to create unique identifiers for people without using facial recognition.
Reinharz added that “having an impartial, hopefully, unbiased system, to help with decision making offers a lot of potential for society because as humans we are so self-interested and so biased even if we are well-intentioned,” providing a more optimistic view that facial recognition could be perfected in the future.
Quang Trinh, Professional Services Manager of Axis Communications, discussed one of the biggest challenges that comes with AI, admitting that “right now, AI is pretty much a wild, wild west.” He explained there are no standards on how data is collected, and he is working with SIA to shape them. Trinh added that it’s important for different groups to work together to determine how best to use AI because it will be implemented much more in the future.
According to Trinh, research shows that “by the end of this decade, the AI global industry is going to reach $1 trillion.” While some of that will be for security applications, the rest will be used for business opportunities.
To learn more about AI, check out SIA’s webinar by clicking here.